Jim Finnerty <jfinnert@amazon.com> writes:
> What were the possible failure scenarios that throwing a fatal error was
> intended to avoid, i.e. what sort of "hooking into" is the comment below
> referring to that was regarded as a fate worse than death?
The point is that if the extension is marking the variable as
PGC_POSTMASTER, it's presumably relying on that variable having the same
value in every process. It might be using it as the size of an array in
shared memory, say. If some processes have a different value, that could
end in a memory stomp, or some other crash that's substantially less clean
than a FATAL exit.
regards, tom lane
